The present embodiments relate to a method for discouraging unsolicited bulk email (spam) and, more particularly, to a method for discouraging spam using a protocol-based technique which imposes a financial penalty on the sender for email items rejected by their recipients.
Electronic mail (email) is now one of the primary ways people and organizations communicate with one another, with billions of emails being sent daily. Email support is offered by all online services and Internet Service Providers (ISPs). Email communication is extremely rapid. Email generally arrives at its destination, anywhere in the world, in minutes. More and more, businesses are relying on email to support commercial activities. For individuals, email has to a large extent replaced writing letters when communicating with friends and family.
With the rise of email has come a concurrent rise in the problem of spam. Spam, otherwise known as unsolicited bulk email (UBE) or unsolicited commercial email (UCE), is in the most simplistic terms electronic junk mail. Since sending an email item is essentially free, spammers are able to send hundreds and thousands of unsolicited emails. Indeed, many spammers do not necessarily ensure that the email addresses are valid. These spammers generate lists of addresses, based on general knowledge of common names and of email providers. When the spammer receives a response from a particular addressee the email address is known to be valid.
It is currently estimated that spam accounts for approximately 40% of all email traffic, with over 12 billion items being transmitted daily. Spam has serious financial implications, as well as personal ones. Internet providers are required to implement costly anti-spam measures, consequently raising the costs of Internet services. Companies lose work-hours, as employees waste significant amounts of time dealing with spam at the workplace. Additionally, spam may contain content seen as offensive by the recipient, such as pornography, or may be an attempt to lure the recipient into a scam. It is estimated that U.S. corporations lost $8.9 billion dollars in 2002 due to spam, and that the cost to non-corporate emailers was $255 million.
A wide variety of measures are currently available to combat spam, and many others have been proposed. Current anti-spam techniques include many types of filters, which attempt to identify whether a given email item is legitimate or should be rejected as spam. Spam filters include:
1. Domain-name filtering—A list is maintained containing the email addresses or domain name of known spammers. Email from senders on the list is rejected. Domain-name filtering is performed at a number of levels. At the email client level, users may maintain a blacklist. When a user receives a spam email, the user enters the spammers email address into the blacklist, either in his email client or his email service. Email coming from that address in the future is rejected. Email services monitor bulk emailings to their users, and attempt to distinguish between legitimate emailers and spammers. A number of public lists of known spammers are available. However, domain-name filtering has proven ineffective, since spammers generally change domain names frequently and are often able to hide their identity.
2. Content-based filtering—The content of an email item is examined to determine whether it is characteristic of spam, for example solicitation to view a pornographic site. Content filtering may be performed at many stages of email transmission, including by the email client, by the email service provider, or by a proxy email filtering service. Spammers are aware of content filtering, and take steps to ensure that the email content is not identifiably spam, for example by using random groupings of letters in the title or deliberately misspelled words.
3. Anti-spam regulations—The U.S. and other countries are beginning to introduce regulations and legislation intended to counter spam. These regulations are considered by many to be a mixed blessing, as they may hinder legitimate as well as illegitimate emailers. Additionally, spammers need only move their operations to an unregulated country to avoid prosecution.
Most of these anti-spam measures place the cost of avoiding spam on the recipient, who must proactively invest time and resources to ensure that he is not troubled by spam that is sent to his email address. A consensus is arising that the way to combat spam is to shift the cost of spam onto the spammers, by requiring email senders to invest a small amount of some resource for each email item sent. In “Stamps vs. Spam” (http://fare.tunes.org/articles/stamps_vs_spam.html-container1069), François-René Rideau discusses a number of proposed approaches.
Microsoft Corp. is promoting an approach known as the Penny Black method. The Penny Black method requires the sender to devote a small amount of computer resources for each email item sent. The computer performs a computation which takes several seconds (for example, producing a binary sequence) prior to sending each email item. For an individual user with ordinary amounts of email the “cost” of sending emails will be insignificant. However, bulk emailers will be obliged to devote large amounts of computer resources to sending emails. It is believed that this cost will discourage spammers, who will be reluctant to purchase and maintain large computer systems in order to send large quantities of unsolicited email. The Penny Black system has two disadvantages. Firstly, it places a large computational burden on legitimate bulk emailers, not just on spammers. Secondly, the economics of spam may be such that it is worthwhile for a spammer to make a one-time capital investment in a computer system that can then be dedicated to sending out spam.
A second proposal is a payment in human resources. The sender would be required to perform a small task for each email item sent. For example, the sender could be instructed by an automated reply from the recipient's software to connect to a web page and to perform a simple task known to be currently impossible for a computer to fulfill (a Turing test). However, payment in human resources is problematic for all legitimate senders of email, regardless of scale. Simplicity, rapidity, and ease of use are considered the great benefits of email. All of these benefits will be lost if the system requires the user to perform unnecessary and complex tasks. Additionally, the current solution may prevent elderly or disabled users from sending email, if they are not able to perform the required task.
A third proposed form of payment is monetary. The sender is required to pay a monetary charge for each email. A small monetary payment would be required for each sent email, which may or may not be refunded to the user if the recipient accepts the email. The charge should be small enough so as not to be a burden on individual and small-scale emailers. Spammers, however, will be financially unable to send out the massive amounts of email they are currently producing. The financially based methods in the related art require a large amount of processing by both users and email service providers, in order to manage an ongoing series of financial transactions. Many provide only a localized-solution for a number of users, which causes additional difficulties by requiring non-standard handling of email to these users. A number of financially based methods for discouraging spam are presented below.
Rideau describes a payment method in which each email item includes a certified check, to be drawn from the sender's account by the recipient or a receipt referring to a payment already done to the recipient's postage account. The recipient's postage software will then bank the check or ensure that the receipt is valid before accepting the email for actual processing.
Since email services are currently free, implementing a monetary charge for email requires establishing the necessary mechanisms to permit charging for email. Spam-proof email accounts are currently being offered to users by proxy email services such as CruelMail, LLC. The user sets a fee, and provides a list of email addresses that aren't subject to delivery fee. All other emailers must pay the required fee or their email is rejected. CruelMail collects the fee for the user filters out unwanted email. The user has the option of monitoring his rejected email, and manually retrieving email items he is interested in viewing, even if they have not paid the fee. CruelMail, and like services, are a local, not an encompassing solution for spam, and moreover, one that is paid for by the recipient. Moreover, the recipient may lose more than the fee paid to the anti-spam service. Email senders are unaccustomed to sending payments to their recipients, and are unlikely to cooperate. Friendly senders will often not want the bother of making even a symbolic payment to the recipient for the privilege of sending an email.
An alternative approach is offered by Goodmail Systems™, which offers an email stamping system. Accredited volume mailers purchase stamps, which can be attached to outgoing email. Participating Internet Service Providers (ISPs) install a stamp-filtering gateway, which clears email bearing a stamp and sends other email via a spam filter. Low-volume emailers may be offered no-cost stamps. However, this solution is also problematic because it applies only to the clients of ISPs which have installed a stamp-filtering gateway, and causes their email handling to be performed in a non-standard way. The current method does not eliminate spam; it only offers bulk emailers a way to bypass a spam filter. This bypass may be ignored by spammers who will continue to take their chances with the spam filter, as well as by bulk emailers who prefer to continue sending unstamped email via the ISPs spam filter.
Various payment schemes have been proposed for charging senders for emails. One such payment scheme is presented in U.S. Pat. No. 6,587,550 by Council et al. An incoming email is checked against a list of authorized parties (i.e. a whitelist). If the sending party is not on the list, an electronic billing agreement is emailed to the sending party indicating a fee that will be charged to the sending party in return for the message being provided to the intended receiving party. After the sender pays the fee, the email is delivered to the recipient. A second payment scheme is given in U.S. Pat. No. 6,697,462 by Raymond. Raymond's anti-spam technique requires senders to post a bond to accompany the communication. The bond is to be forfeited if the communication is rejected or deemed undesirable by the recipient.
The monetary anti-spam schemes proposed so far all require integrating sophisticated billing mechanisms into email services. The billing mechanism would be involved in virtually every email transaction, and would potentially be required to handle millions of micro-payments daily. Service providers may need to make costly investment in capital equipment and billing software, which costs will be passed on to their users. In order to limit the number of financial transactions, users will be required to maintain close control of their whitelists. Additionally, problems that arise during the payment process may require multiple communications between sender, recipient, and service provider in order to resolve financial issues, thus increasing the complexity and time required for email communications rather than reducing them.
There is thus a widely recognized need for, and it would be highly advantageous to have, an anti-spam method of email transmission, devoid of the above limitations.